On election eve, Arlene Battishill, a retired history professor and political analyst known as “Dr. Arlene Unfiltered” to her social media followers, picked up a bottle of champagne, confident in the data across different states that showed a potential Kamala Harris win.
But a different narrative had emerged on Polymarket, a decentralized prediction platform founded in 2020, built on blockchain technology and banned in the U.S. It had consistently shown Trump leading since his September debate with Harris, and it ultimately proved more accurate than conventional polling methods that showed the race in a dead heat.
“This is the third consecutive presidential election in which [Donald] Trump was underestimated by polls,” said Harry Crane, a professor of statistics at Rutgers University.
